METHODS FOR IMPROVING SOFTWARE CORRECTNESS AND SECURITY: MODERN COMPILER FUNCTIONALITY AND OTHER APPROACHES

Abstract

In the context of the rapid adoption of digital technologies, software plays a crucial role in the functioning of critically important domains – ranging from financial and medical systems to infrastructure solutions, security tools, and industrial automation. Accordingly, the demands on the quality, reliability, and security of software systems continue to grow. One of the key factors shaping these characteristics is software correctness, defined as the compliance of a program’s behavior with its formal or informal specification in all permissible situations.Ensuring correctness is a complex task that requires both deep understanding of the application domain and careful technical modeling. Common sources of deviations include logical, structural, or conceptual errors, which may destabilize system operation, lead to data loss or corruption, and in some cases – introduce serious security vulnerabilities.Given these risks, tools for early detection and prevention of defects become particularly significant. Among them, modern compilers stand out for their expanded functionality, which goes beyond traditional syntactic analysis. They are now capable of performing advanced semantic checks – including completeness of condition handling, validity of parameter values, index correctness, and more. Static code analysis tools also play an essential role, allowing assessment of program behavior without actual execution.Despite the availability of such powerful tools, current engineering practice often prioritizes rapid product deployment. As a result, correctness concerns are frequently sidelined – especially in startup environments or under agile methodologies. However, in high-risk sectors such as healthcare, energy, or security, even minor behavioral deviations can have catastrophic consequences.This article presents an analysis of both classical and contemporary publications devoted to improving software quality. Some of them offer structured methods for ensuring correctness, while others propose conceptual approaches to designing software that minimizes the likelihood of errors. Particular attention is paid to mechanisms for correctness verification at the compilation stage.